Method and system for application migration in a cloud

• US 8,751,627 B2
• Filed: 05/05/2009
• Issued: 06/10/2014
• Est. Priority Date: 05/05/2009
• Status: Active Grant

• Current Assignee: Accenture Global Services Limited (Accenture PLC)
• Original Assignee: Accenture Global Services Limited (Accenture PLC)
• Inventors: Orban, Dan, Liu, Huan
• Primary Examiner(s): Gillis, Brian J
• Application Number:
  US12/435,596
• Publication Number:
  US 20100287263A1
• Time in Patent Office: 1,862 Days
• Field of Search: 709/221, 709/223, 709/224, 709/238, 709/249
• US Class Current: 709/224
• CPC Class Codes: G06F 11/203 using migration
• CPC Class Codes: G06F 9/5088 involving task migration
• CPC Class Codes: H04L 63/1408 by monitoring network traffic monitoring network traffic per se H04L43/00
• CPC Class Codes: H04L 63/1441 Countermeasures against malicious traffic countermeasures against attacks on cryptographic mechanisms H04L9/002
• CPC Class Codes: H04L 63/1458 Denial of Service
• CPC Class Codes: H04L 65/80 Responding to QoS
• CPC Class Codes: H04L 67/10 in which an application is distributed across nodes in the network software deployment G06F8/60; multiprogramming arrangements G06F9/46

Abstract

A method and system for managing an application in a cloud data center by monitoring the bandwidth of a subnet of which the primary operating instance of an application is a member. If a severe deterioration in the bandwidth caused by an over consumption of the subnet resources from other subnet constituents is detected, a suitable secondary instance in an alternate, uncompromised subnet is located and primary operation of the application is transferred from the former primary operating instance to the secondary instance. The secondary instance can be pre-launched or dynamically requisitioned in the cloud data center.

Claims

1. A method for application migration from a first computing environment to a second computing environment, comprising:

receiving a distress signal as a plurality of UDP datagrams indicating a degraded quality of service of a first instance of an application executing in a first computing environment, the indication of the degraded quality of service comprising a first bandwidth of the first instance of the application being determined to be below a first threshold;

in response to receiving the distress signal, dynamically provisioning a second application executing in a second computing environment based on detecting a degraded quality of service of the first application, the second computing environment having a corresponding second bandwidth determined to be above a second threshold; and

transferring a primary operation from the first application to the second application executing in the second computing environment.

2. The method according to claim 1, wherein the distress signal is generated in response to detecting bandwidth starvation in a networking device providing network routing service to a subnet of the plurality of subnets, the subnet comprising a plurality of computing environments, including the first computing environment.

3. The method according to claim 2, wherein detecting bandwidth starvation comprises sending a plurality of probes into the subnet.

4. The method according to claim 3, wherein a probe of the plurality of probes is a packet pair, the packet pair comprising a first packet and a second packet.

5. The method according to claim 4, wherein detecting bandwidth starvation comprises:

sending a packet pair to a destination outside of the subnet; and

measuring a time delay between receiving a response corresponding to the first packet of the packet pair and a response corresponding to the second packet of the packet pair.

6. The method according to claim 3, wherein sending a plurality of probes comprises sending a plurality of single probe packets.

7. The method according to claim 6, wherein detecting bandwidth starvation comprises:

measuring a queueing delay corresponding to the plurality of single probe packets in the networking device.

8. The method according to claim 2, wherein the second computing environment is comprised in a different subnet than the subnet comprising the first computing environment.

9. The method according to claim 1, wherein transferring the primary operation comprises:

re-assigning a second IP address corresponding to the second application with a first IP address corresponding to the first application.

10. The method according to claim 1, wherein transferring the primary operation comprises:

re-programming a DNS resolution of a domain name corresponding to the first application to correspond to the second application.

11. A method for managing an application to avoid low bandwidth, comprising:

detecting a deterioration of a first bandwidth of a first instance of an application executing on a first computing environment disposed in a first subnet of a data network, wherein the first instance of the application is a primary operating instance of the application;

receiving a distress signal as a plurality of UDP datagrams from the first instance of the application indicating a deterioration of the first bandwidth;

dynamically launching a stand-by instance of the application on a second computing environment in a second subnet of the data network in response to detecting deterioration of the first bandwidth of the first instance;

measuring a second bandwidth available to the second subnet of the data network,

comparing the second bandwidth to a second pre-determined threshold; and

initiating an application migration from the first computing environment to the second computing environment if the second bandwidth is greater than the second pre-determined threshold.

12. The method according to claim 11, wherein measuring the stand-by instance of the application comprises measuring a bandwidth of a pre-launched stand-by instance for the first instance.

13. The method according to claim 11, wherein measuring the second bandwidth comprises:

measuring a bandwidth available to the stand-by instance of the application.

14. The method according to claim 11, the method further comprising:

de-activating the first instance of the application as the primary operating instance of the application;

activating the stand-by instance of the application as the primary operating instance of the application; and

launching a third instance of the application executing on a third computing environment in a third subnet of the data network, the third computing environment comprising a new stand-by instance for the stand-by instance of the application,

wherein the de-activating, the activating and the launching are performed subsequent to the initiating the application migration from the first computing environment to the second computing environment.

15. The method according to claim 14, wherein changing a DNS translation of a domain name corresponding to the application comprises changing the domain name corresponding to the application in a plurality of name servers from translating to an IP address of the first instance of the application to an IP address of the stand-by instance of the application.

16. The method according to claim 11, the method further comprising:

de-activating the first instance of the application as the primary operating instance of the application;

activating the stand-by instance of the application as the primary operating instance of the application; and

designating a pre-launched third instance of the application as a new stand-by instance for the stand-by instance of the application, the pre-launched third instance of the application executing on a third computing environment in a third subnet of the data network,

wherein the de-activating, the activating and the designating are performed subsequent to the initiating the application migration from the first computing environment to the second computing environment.

17. The method according to claim 11, wherein detecting a deterioration of a first bandwidth comprises:

estimating the first bandwidth;

comparing the first bandwidth with a first pre-determined threshold; and

determining the first bandwidth is lower than the first pre-determined threshold.

18. The method according to claim 17, wherein estimating the first bandwidth comprises:

sending a plurality of marked data packets to the first instance of the application;

receiving a plurality of return packets corresponding to the plurality of marked data packets, wherein the plurality of return packets have a plurality of arrival times; and

using differences between the plurality of arrival times to estimate the first bandwidth.

19. The method according to claim 11, further comprising:

filtering a plurality of communications from the first instance of the application.

20. The method according to claim 11, further comprising:

launching a plurality of stand-by instances of the application in a plurality of computing environments comprised in a plurality of subnets of the data network;

estimating a plurality of bandwidths available to the plurality of computing environments;

evaluating the plurality of bandwidths relative to the second pre-determined threshold; and

migrating the application from the first computing environment to a third computing environment comprised in the plurality of computing environments if the third computing environment has a bandwidth greater than first pre-determined threshold,

wherein the launching, estimating, evaluating and migrating are performed if the second bandwidth is not greater than the first bandwidth.

21. The method according to claim 11, wherein the second pre-determined threshold is greater than the first bandwidth.

22. The method according to claim 11, wherein the measuring a second bandwidth is repeated periodically.

23. The method according to claim 11, wherein initiating an application migration from the first computing environment to the second computing environment comprises switching a static IP address of the second computing environment to the static IP address of the first computing environment.

24. The method according to claim 11, wherein initiating an application migration from the first computing environment to the second computing environment comprises changing a DNS translation of a domain name corresponding to the application.

25. An application management system in a cloud data center, the system comprising:

a data center network comprising a plurality of subnets, the plurality of subnets comprising a plurality of computing environments, the plurality of computing environment communicatively coupled by a plurality of networking devices corresponding to the plurality of subnets;

a first computing environment of the plurality of computing environments, the first computing environment comprised in a first subnet of the plurality of subnets communicatively coupled to a first networking device having a first bandwidth;

a first instance of an application, the first instance of the application being executed on the first computing environment and comprising a primary operation of the application; and

a monitoring agent for managing the application, the monitoring agent being configured to receive distress signals from the first instance of the application,

wherein the first instance of the application transmits a distress signal comprising a plurality of UDP datagrams to the monitoring agent when the first instance of the application experiences a bandwidth below a pre-determined threshold,

wherein when the first instance of the application experiences a bandwidth below a first pre-determined threshold, the monitoring agent transfers the primary operation of the application to a dynamically-launched second instance of the application executed on a second computing environment of the plurality of computing environments comprised in a second subnet of the plurality of subnets, the second computing environment having a corresponding bandwidth above a second pre-determined threshold.

26. The system according to claim 25 wherein at least one of the first computing environment and the second computing environment comprises a server.

27. The system according to claim 26, wherein at least one of the first computing environment and the second computing environment comprises a virtual machine.

28. The system according to claim 25, wherein a second subnet comprises a different subnet than the first subnet.

29. The system according to claim 25, wherein the dynamically launched second instance of the application comprises a stand-by instance of the application.

30. The system according to claim 29, wherein the dynamically launched second instance of the application is dynamically launched in response to when the first instance of the application experiences a bandwidth below a pre-determined threshold, and in advance of the transference of the primary operation of the application.

31. The system according to claim 25, wherein the monitoring agent is executed on a third computing environment within the data center network, wherein the third computing environment is not disposed in the first subnet.

32. The system according to claim 25, wherein the monitoring agent is implemented as a load balancer.

33. The system according to claim 25, wherein the monitoring agent is executed on the second computing environment.

34. The system of claim 25, wherein the monitoring agent is executed on a fourth computing environment externally disposed with respect to the data center network, and the fourth computing environment is communicatively coupled to the data center network.

35. The system of claim 25, wherein the bandwidth below a pre-determined threshold comprises a bandwidth in a subnet reduced from over-consumption by constituents of a subnet.

36. The system of claim 25, wherein the bandwidth below a pre-determined threshold comprises a bandwidth in a subnet reduced by a denial of service attack.